Project Summary Delayed language development is one of parents' most common concerns expressed during well-child visits at 24 months of age. This high rate of concern is reflected in research findings that approximately 15% of toddlers experience delayed expressive language development (and are labeled as ?late talkers?). This group of children has highly variable long-term outcomes, with some children experiencing persistent language problems, and others developing typical language skills. While effective early interventions exist, a critical barrier to improving the health and long-term outcomes of these children is a lack of reliable diagnostic predictors of long-term outcomes. Current diagnostic assessments cannot differentiate between those children who will spontaneously recover and those who need language intervention services. Without improved predictors, pediatricians and clinicians cannot allocate limited intervention resources to those children most in need. Typical diagnostic measures used both clinically and in research include behavioral assessments of children's broad expressive and receptive language skills. However, these gross measures of linguistic development alone have poor predictive validity. This project proposes to use more precise, fine-grained measures of two psycholinguistic skills in order to characterize language profiles in late talkers: phonological development, and semantic sensitivity. This will be accomplished by utilizing a multi-method, multi-informant approach. Children's sensitivity to semantic associations will be measured receptively using an intermodal preferential looking paradigm (IPLP) eye tracking task (Aim 1a). Semantic development will also be assessed through modeling the structure of children's expressive vocabulary using network analysis of their early, parent-reported vocabulary (Aim 1b). Children's receptive phonological development will be measured using a second IPLP eye tracking task to assess their sensitivity to mispronunciation in early-acquired words (Aim 2a). Additionally, the phonological structure of children's early expressive vocabulary will be modeled using network analytic techniques (Aim 2b). Finally, an exploratory aim will use Latent Profile Analysis to test the extent to which there are distinct profiles of performance across these four tasks within late talkers (Aim 3). Understanding how these two psycholinguistic skills are impacted in late talkers, and the extent to which distinct profiles of skills exists in these children, will have a positive translational impact by elucidating areas for intervention in this population and aiding in determining which children are most likely in need of intervention services.